Skip to main content

Browse the archive

https://github.com/E3SM-Project/E3SM
06 April 2024, 03:10:57 UTC
sort by:
Revision Author Date Message Commit Date
90893f5 Jinyun Tang 09 September 2019, 17:15:08 UTC update betr branch ensure betr is update to date. 09 September 2019, 17:15:08 UTC
049c0bd Jinyun Tang 21 March 2019, 15:04:12 UTC zero out initial finudation A zero value is implemented for using finudation in betr. 21 March 2019, 15:04:12 UTC
8cc1be5 Jinyun Tang 19 March 2019, 17:30:14 UTC some cleanup and enhancement for exit spinup Also updated sbetr with most recent tag. 19 March 2019, 17:30:14 UTC
2d07d3b Jinyun Tang 10 March 2019, 22:33:23 UTC bug fix for v1eca Added more diagnostics for v1eca and also made sure agnpp is on the output list. 10 March 2019, 22:33:23 UTC
2708e09 Jinyun Tang 05 March 2019, 00:57:47 UTC bug fix for v1eca Fixed a number of parameter transfer bugs between betr and elm for v1eca. 05 March 2019, 00:57:47 UTC
b16d23b Jinyun Tang 28 February 2019, 22:47:41 UTC bug fix and add ecosys N fix to output Fixed a duplicated calculation of some carbon fluxes and put ecosys N fix to history file. 28 February 2019, 22:47:41 UTC
5fabdc4 Jinyun Tang 24 February 2019, 03:05:30 UTC bug fix for betr-v1eca Fixed the C-N mass balance error for global simulations. 24 February 2019, 03:05:30 UTC
0cbff07 Jinyun Tang 19 February 2019, 19:51:26 UTC bug fix for v1eca-betr coupling Now v1eca passed test with full bgc at a single point. 19 February 2019, 19:51:26 UTC
fff4957 Jinyun Tang 14 February 2019, 00:12:12 UTC bug fix for v1eca-elm coupling Fixed some calling sequence error and ensured v1eca is working at a single site. 14 February 2019, 00:12:12 UTC
33d1feb Jinyun Tang 11 February 2019, 01:49:17 UTC bug fix for betr v1eca Fixed a bug in calling sequence that caused carbon mass balance error. 11 February 2019, 01:51:24 UTC
b3500ca Jinyun Tang 10 February 2019, 03:02:02 UTC a number of initialization bug fixes Fixed a number number of bugs related to initialization and coupling with v1eca. 10 February 2019, 03:02:02 UTC
deec736 Jinyun Tang 07 February 2019, 04:55:41 UTC updated P flux variables for v1eca dry run Now exported net P mineralization to support v1eca. 07 February 2019, 04:55:41 UTC
c48f5ba Jinyun Tang 07 February 2019, 03:58:17 UTC bug fix for type0 bgc Fixed a number of declaration bugs. 07 February 2019, 03:58:17 UTC
d4e478d Jinyun Tang 07 February 2019, 02:57:19 UTC use module update for type0 betr bgc This ensures states update are using betr specific modules. 07 February 2019, 02:57:19 UTC
6e4ae65 Jinyun Tang 06 February 2019, 23:58:08 UTC bug fix for root autotrophic respiration summary This is a fix for betr coupling. 06 February 2019, 23:58:08 UTC
0068529 Jinyun Tang 06 February 2019, 22:51:50 UTC updated betr coupling for v1eca dry run test Now the model includes further soil bgc variables being exchanged between elm and sbetr. It is ready for the dry run test. 06 February 2019, 22:51:50 UTC
eb46532 Jinyun Tang 06 February 2019, 04:49:17 UTC bug fix for elm-betr coupling Fixed a number of declaration bugs related to betr interface. 06 February 2019, 04:50:06 UTC
562db26 Jinyun Tang 05 February 2019, 20:22:38 UTC updated elm-betr interface Added further coupling components between elm and betr for v1eca. 05 February 2019, 20:22:38 UTC
9a11817 Jinyun Tang 05 February 2019, 00:17:38 UTC bug fix for betr bgc interface Fixed an number declaration bugs. 05 February 2019, 00:17:38 UTC
1c7e227 Jinyun Tang 04 February 2019, 23:21:09 UTC update interface for betr-elm coupling. Separated out type0 and type1 elm-bgc coupling strategies. 04 February 2019, 23:21:09 UTC
b762f9e Jinyun Tang 31 January 2019, 04:17:42 UTC bug fix of a do loop Removed an unncessary enddo clause. 31 January 2019, 04:17:42 UTC
496b33b Jinyun Tang 31 January 2019, 03:20:09 UTC implemented summary_bgc_cascade This moves elm soil bgc flux summary into separate subroutines, which allows them to be turned off when external soil bgc is is on. 31 January 2019, 03:24:37 UTC
a325baf Jinyun Tang 31 January 2019, 00:04:33 UTC cleared summary subroutine in CNCarbonFluxType This is the first part of output clear up. 31 January 2019, 00:04:33 UTC
5912cac Jinyun Tang 30 January 2019, 22:51:52 UTC prepare to soil bgc io refactoring This added code to do elm soil bgc flux summary. It prepares for turn off default bgc flux summary when betr bgc is on. 30 January 2019, 23:15:31 UTC
061c6f8 Jinyun Tang 30 January 2019, 21:06:26 UTC bug fix after interface update for type1_bgc Fixed an interface bug and syncrhonized with the most recent betr branch. 30 January 2019, 21:06:26 UTC
876e2b1 Jinyun Tang 30 January 2019, 18:47:57 UTC further interface update for betr type1_bgc Updated the list of variables passed in to betr. 30 January 2019, 18:47:57 UTC
67ca335 Jinyun Tang 27 January 2019, 05:39:07 UTC enhanced betr-elm coupling for type1_bgc Fixed a number build bugs and is ready to implement the type1_bgc within betr. 27 January 2019, 05:39:07 UTC
57f44ce Jinyun Tang 27 January 2019, 05:19:18 UTC interface update for betr-elm coupling Enhanced compatability with type1_bgc coupling. No change to default elm bgc. 27 January 2019, 05:19:18 UTC
1ffea15 Jinyun Tang 25 January 2019, 21:23:13 UTC bug fix for passing kinetic parameters to betr Fixed a number of declaration bugs and successfully did a type0_bgc test. 25 January 2019, 21:23:13 UTC
df25026 Jinyun Tang 25 January 2019, 19:27:02 UTC add interface to pass out eca competition parameters This prepares to send kinetic parameters to sbetr for plant soil-coupling. 25 January 2019, 19:31:59 UTC
a7e49ed Jinyun Tang 24 January 2019, 22:13:48 UTC compiling bug fix for updated elm-betr coupling It fixed some declaration bugs in the new subroutine for innudation calculation. 24 January 2019, 22:13:48 UTC
f90fb86 Jinyun Tang 23 January 2019, 23:25:27 UTC update interface for elm-betr type1_bgc Updated the calling sequence for type1_bgc coupling between elm and betr. 23 January 2019, 23:25:27 UTC
d88eaee Jinyun Tang 22 January 2019, 17:16:59 UTC bug fix for elm-betr coupling Fixed some declaration bug. 22 January 2019, 17:16:59 UTC
517eb67 Jinyun Tang 21 January 2019, 04:51:16 UTC update elm-betr coupling for type0_bgc This enables a no oxygen bgc test with elm-betr interface. 21 January 2019, 04:51:16 UTC
c060393 Jinyun Tang 19 January 2019, 01:12:23 UTC bug fix for elm-betr coupling Updated the betr-elm interface. Preparing for testing v0 coupling. 19 January 2019, 01:12:23 UTC
3b648da Jinyun Tang 17 January 2019, 21:21:50 UTC update betr related code for elm-betr coupling Now added one functionality to do lightweight test of using elm-eca within betr. 17 January 2019, 21:21:50 UTC
d18d1fe Jinyun Tang 07 January 2019, 21:55:23 UTC remove obsolete files and prepare for sbetr update Removed some redudant files and prepare for merging with the up to date sbetr. 07 January 2019, 21:55:23 UTC
0397d44 Jinyun Tang 07 January 2019, 21:28:42 UTC implemented the multi-substarte flux-limiter to plant This resolves an issue of negative little fluxes. It also allows much more accurate precision control for vegetation C N and P state variables. This change is not bit for bit. 07 January 2019, 21:28:42 UTC
80156d9 Jinyun Tang 05 January 2019, 02:19:37 UTC apply the flux limiter for vegetation fluxes This attempts to fix the excessive carbon and nutrient fluxes that are not well constrained during phenological allocation. This may remove the necessity of precision control in the vegetation components. 05 January 2019, 02:19:37 UTC
d18707b Mark Taylor 24 December 2018, 15:07:31 UTC Merge branch 'ambrad/ambrad/homme/scalable-init-phase3' (PR #2601) This PR has three parts. In the first commit, deallocation of GridVertex and GridEdge is moved closer in time to their allocation for efficiency of memory use. In the second commit, the space filling curve initialization is made scalable, as explained in more detail in that commit message. This work complements PR #2579, which focused on making GridVertex and GridEdge scalable. Edit: The commit message should say that O(log ne) memory is used; it incorrectly says O(ne). In the third commit, the SFC part of initializing a mesh-file-based grid for RRM is modified to use the new SFC functions and so is made to be scalable. [BFB] 24 December 2018, 15:07:31 UTC
ab64fc3 Mark Taylor 24 December 2018, 15:06:20 UTC Merge branch 'oksanaguba/nonhydro/homme/theta-ftype4' (PR #2578) This PR adds ftype 4 (dribbling tendencies at each timestep) to standalone HOMME theta models. 24 December 2018, 15:06:20 UTC
efcda7b Oksana Guba 22 December 2018, 16:53:51 UTC changes after review 22 December 2018, 16:53:51 UTC
37e51c6 James Foucar 20 December 2018, 20:34:43 UTC Merge branch 'worleyph/cime/flush_fix' into master (PR #2670) replace flush with shr_sys_flush In a recent PR, a call to flush was introduced into CIME, but the CIME style is to use shr_sys_flush. This PR corrects this discrepancy. [BFB] * origin/worleyph/cime/flush_fix: replace flush with shr_sys_flush Update CIME to ESMCI cime5.7.5 (PR #2661) 20 December 2018, 20:35:40 UTC
b9a9e91 Patrick Worley 19 December 2018, 16:31:31 UTC replace flush with shr_sys_flush In a recent PR, a call to flush was introduced into CIME, but the CIME style is to use shr_sys_flush. This PR corrects this discrepancy. [BFB] 19 December 2018, 16:31:31 UTC
584a2f8 James Foucar 12 December 2018, 17:02:24 UTC Merge branch 'worleyph/cime/taskmap_control' into master (PR #2622) Control ouput of MPI task to compute node mapping data A recent PR fixed the output of component-specific MPI task to compute node mapping for ATM and LND, and added it for ROF, OCN, ICE, and GLC. The utility of this output for each component is unclear, and a namelist variable to enable or disable it is added here. This variable also allows setting the format of the output. All component model task-to-node mapping output is controlled by this single cime_pes namelist variable info_taskmap_comp, where setting this to 0 disables output, setting it to 1 uses the more compact format, and setting it to 2 uses the more verbose format. Note that a separate component-only PR will be needed for the components to be aware of info_taskmap_comp and use this as described above, but this additional code has been used to test this PR. For symmetry, a cime_pes namelist variable info_taskmap_model was also added to control the task-to-node mapping output for the full coupled model, with the same options (0, 1, and 2). The default here is 2 (preserving the current behavior). Also for symmetry, output of task-to-node mapping data was added for the CPL component (and is controlled by info_taskmap_comp the same as for the other component models). Finally, the implementation for the whole-model task-to-node mapping data output was modified to better reflect the (currently incomplete?) support for running the model with multiple drivers. The two new namelist parameters were added to env_run.xml, and should be set from there. [BFB] [NML] * origin/worleyph/cime/taskmap_control: Change write format specification to match existing style Add control of taskmap output to env_run.xml Control ouput of MPI task to compute node mapping data 18 December 2018, 17:32:47 UTC
97f3c7a James Foucar 12 December 2018, 18:57:50 UTC Update CIME to ESMCI cime5.7.5 (PR #2661) Update CIME to ESMCI cime5.7.5 Squash merge of jgfouca/branch-for-to-acme-2018-12-11 Features: * Merge latest nuopc-cmeps development into cime master * Big improvement to robustness of CIME's XML handling Bug Fixes: * Fix typo in setting strm_datdir for CLM_QIAN_WISO.TPQW * FUNIT test: Fix path to run_tests.py if get_test_spec_dir is overridden * Allow CHARGE_ACCOUNT to use subprojects and default to command line given project * Remove redundant check of mpi-tag values * Update lnd2glc_averaged_now in the driver loop * Fix regex matches in hist_utils: get_extension * Fix new pylint errs * fix PFS generate issue [BFB] 18 December 2018, 17:28:52 UTC
da54334 Balwinder Singh 14 December 2018, 21:18:58 UTC Merge branch 'singhbalwinder/atm/pgn-rename-cld-file' (PR #2664) Changes file name of cloud file so that bless script can copy it [BFB] - Bit-For-Bit 14 December 2018, 21:18:58 UTC
dafe289 James Foucar 14 December 2018, 20:50:54 UTC Merge branch 'pwburge/cime/edit-config_machines.xml-to-add-overload' into master (PR #2660) Allow cpu overloading for MPI for melvin [BFB] * origin/pwburge/cime/edit-config_machines.xml-to-add-overload: Allow cpu overloading for MPI for melvin 14 December 2018, 20:51:28 UTC
eaf53f9 Balwinder Singh 14 December 2018, 20:02:12 UTC Changes file name of cloud file so that bless script can copy it [BFB] - Bit-For-Bit 14 December 2018, 20:02:12 UTC
a4ac51d Balwinder Singh 13 December 2018, 17:48:02 UTC Merge branch 'singhbalwinder/atm/pgn-gen-cloud-reorder' (PR #2662) Reorders code to allow writing PGN cloud in the run phase PGN was generating its probability cloud only in "generate" phase. This cloud should be generated in the "run" phase so that baseline blessing script can copy this file over when we bless baselines. [BFB] - Bit-For-Bit 13 December 2018, 17:48:02 UTC
f9edfca Balwinder Singh 13 December 2018, 15:57:48 UTC Uses capital global vars and logger.debug [BFB] - Bit-For-Bit 13 December 2018, 15:57:48 UTC
365701c Balwinder Singh 12 December 2018, 19:18:59 UTC Reorders code to allow writing PGN cloud in the run phase PGN was generating its probability cloud only in "generate" phase. This cloud should be generated in the "run" phase so that baseline blessing script can copy this file over when we bless baselines. [BFB] - Bit-For-Bit 12 December 2018, 19:18:59 UTC
6e03bc7 Wade Burgess 11 December 2018, 18:15:15 UTC Allow cpu overloading for MPI for melvin [BFB] 11 December 2018, 18:26:27 UTC
9ae414b James Foucar 11 December 2018, 18:06:08 UTC Merge branch 'jgfouca/cime/update_melvin_openmpi' into master (PR #2659) The old version that was being used, 1.8.7, was hanging for unknown reasons on an all_gather that was recently added for computing MPI task to node mappings. [BFB] * jgfouca/cime/update_melvin_openmpi: Update melvin openmpi 11 December 2018, 18:07:07 UTC
b078498 James Foucar 11 December 2018, 18:01:06 UTC Update melvin openmpi The old version that was being used, 1.8.7, was hanging for unknown reasons on an all_gather that was recently added for computing MPI task to node mappings. [BFB] 11 December 2018, 18:01:06 UTC
c16e2c9 Jon Wolfe 10 December 2018, 15:56:54 UTC Merge branch 'jonbob/mosart/update-8th-degree-112018' (PR #2639) Update the MOSART 1/8th degree global parameter file This PR brings in a new MOSART 1/8th-degree global parameter file that corrects the unrealistic values in the uh (Manning's roughness coefficient for overland flow routing) field. The new file is called MOSART_global_8th_20180211c.nc and has been uploaded to the inputdata directory on anvil. [NML] [non-BFB] only for high-res configurations with active runoff 10 December 2018, 15:56:54 UTC
4df907a Balwinder Singh 06 December 2018, 23:38:35 UTC Merge branch 'singhbalwinder/atm/pgn-tsc-update-lnd-init' (PR #2653) Updates finidat for tsc and pgn test This PR updates finidat files path for tsc and pgn tests. These tests are currently using these files for generating ensemble members. The old files are not working as finidat file now has a new format. See PR #2564 for more details [BFB] - Bit-For-Bit 06 December 2018, 23:38:35 UTC
3e0ad7d Balwinder Singh 06 December 2018, 20:29:31 UTC Updates finidat for tsc and pgn test This PR updates finidat files path for tsc and pgn tests. These tests are currently using these files for generating ensemble members. The old files are not working as finidat file now has a new format. See PR #2564 for more details [BFB] - Bit-For-Bit 06 December 2018, 20:29:31 UTC
a550aa9 James Foucar 06 December 2018, 18:28:12 UTC Merge branch 'worleyph/cime/syslog_upd' into master (PR #2642) Update system-specific job progress monitoring scripts When developing the job progress monitoring script for Summit, some improvements were identified that apply equally to the scripts for the other systems (syslog.). On Summit, background processes ignore kill signals while they are asleep, and job scripts will not complete until spawned background processes are killed or complete normally. Thus, if the job progress monitoring script sleeps for a long time (a user-defined interval), allocation time will be wasted. The workaround was to divide the sleep interval into a loop over 2 minute sleep intervals. This same change of implementation is introduced here for all systems, to make sure that background jobs do not outlive the spawning job scripts. Currently, some system status information is collected twice (with output of the second call overwriting that of the first). A second clean-up is to remove this redundancy. The e3sm log file is copied to the checkpoint directory at the beginning of the run, in an effort to capture the task-to-node mapping. However, this copy needs to be delayed long enough for the mapping information to show up in the log file first. Various methods have been tried to calculate the number of lines in e3sm.log that would indicate that this information has been captured. Yet another approach is implemented here, using empirically determined system-specific and job-specific line counts. Note that we cannot finish the update for Summit until Summit is available for testing again (perhaps Jan. 1, 2019?), but the code that needs to be tested has been added and commented out in the meantime. Finally, because Mira and Cetus are no longer target systems for the E3SM project, the corresponding syslog scripts are deleted. [BFB] * origin/worleyph/cime/syslog_upd: Update logic on when to copy over e3sm.log file for Summit Update system-specific job progress monitoring scripts 06 December 2018, 18:33:39 UTC
d6b5286 Jon Wolfe 03 December 2018, 15:51:54 UTC Merge branch 'mark-petersen/update_mpas_source' (PR #2640) Update mpas-source submodule This update to mpas-source includes: 1. ocean: Update BGC commit hash 2. sea ice: Correction to prevent thermo energy conservation error 3. ocean: fix mixed layer depth for land ice 4. ocean: add variable-resolution GM Bolus capability (4) adds five new ocean config flags, but default behavior for GM remains the same. [NML] [non-BFB] 03 December 2018, 15:51:54 UTC
99f11b3 Jon Wolfe 03 December 2018, 15:48:01 UTC Merge branch 'darincomeau/clm/update_finidat' (PR #2645) Update name of finidat file for cryosphere compsets This PR updates the name of the finidat file used for cryosphere 1850 CMIP6 compsets, which is needed after #2564. This PR also changes (reduces) the default extent of the sea ice initial condition for ice grids oEC60to30v3wLI / oRRS30to10v3wLI. This will change namelists and answers only for tests with the oEC60to30v3wLI and oRRS30to10v3wLI grids. [NML] - only for tests with oEC60to30v3wLI / oRRS30to10v3wLI [non-BFB] - only for tests with oEC60to30v3wLI / oRRS30to10v3wLI 03 December 2018, 15:48:01 UTC
5a6d9ff Oksana Guba 01 December 2018, 06:06:24 UTC disabling tests 01 December 2018, 06:06:24 UTC
7c3e44d Oksana Guba 01 December 2018, 06:02:23 UTC adding dcmip16 TC tests for preqx ahd h theta-l with ftype4 and moist 01 December 2018, 06:02:23 UTC
d3e6cae Gautam Bisht 30 November 2018, 20:39:16 UTC Merge branch 'rgknox/lnd/fates-api' (PR #2618) This changeset bring E3SM compatibility up to the most recent FATES API 6.0.0. This change enables the switching of different plant allocation modules "fates_parteh_mode" in the namelist file. This change also added a call in the interface to accomodate preprocessing some extra information needed to correctly restart the radiation environment. [non-BFB] 30 November 2018, 20:39:16 UTC
169f28a mark-petersen 30 November 2018, 14:48:11 UTC Update mpas-source to fix kappa typos in Registry 30 November 2018, 14:48:11 UTC
6f4de7b Darin Comeau 30 November 2018, 02:12:35 UTC Updating sea ice initial condition extents for oRRS30to10v3wLI grid. 30 November 2018, 02:12:35 UTC
a124876 Darin Comeau 30 November 2018, 02:03:02 UTC Updating name of finidat file for cryosphere compsets. Updating extent of sea ice initial condition. 30 November 2018, 02:03:02 UTC
a14e641 Junqi Yin 29 November 2018, 20:22:03 UTC Merge branch 'bishtgautam/lnd/update-mpp' (PR #2615) Updates the MPP submodule Brings in the latest version of the MPP library in E3SM. [BFB] 29 November 2018, 20:22:03 UTC
94cbaec mark-petersen 29 November 2018, 19:05:07 UTC Fix kappa typos in xml file. 29 November 2018, 19:05:07 UTC
b58d556 Balwinder Singh 28 November 2018, 17:21:35 UTC Merge branch 'singhbalwinder/machines/cascade-lmod' (PR #2629) Modifies lmod path for Cascade machine (PNNL) [BFB] - Bit-For-Bit 28 November 2018, 17:21:35 UTC
4044392 Gautam Bisht 06 November 2018, 02:37:05 UTC Updates the MPP submodule Brings in the latest version of the MPP library in E3SM. [BFB] 28 November 2018, 15:28:24 UTC
dd15fa3 Patrick Worley 28 November 2018, 00:28:51 UTC Update logic on when to copy over e3sm.log file for Summit Logic on when to copy over the e3sm.log into the checkpoint directory has been modified for Summit, similar to the modifications for the other systems. Part of this new logic was then commented out because we can not yet test this on Summit. When Summit is available for user codes, probably after January 1, we will test all of the new code and uncomment it out if it works as expected. [BFB] 28 November 2018, 00:28:51 UTC
e1cc6ad mark-petersen 27 November 2018, 15:15:21 UTC Add GM Bolus ramp to E3SM namelist defaults 27 November 2018, 15:15:21 UTC
b3a0e50 Robert Jacob 26 November 2018, 22:52:07 UTC Merge branch 'worleyph/components/taskmap' (PR #2486) Use shr_taskmap_write in CAM, CLM, Mosart, mpas-albany-landice, mpas-ocean, mpas-seaice to write out all task-to-node mapping information in the new more compact representation, and redirect the output to the associated component log file. a) Output CAM MPI task to compute node mapping using share routine MPI task to compute node mapping information in CAM is output one line per task, and only for the first 256 MPI tasks. A new share routine in CIME (shr_taskmap_write) is available that writes out one line per compute node. Each line contains the compute node name and the list of MPI tasks assigned to that node for a given communicator. This share routine also optionally returns the number of compute nodes and the task-to-node mapping, information that is needed in the internal CAM load balancing. Here the CAM logic in spmdinit (in spmd_utils.F90) is removed and a call to shr_taskmap_write in atm_init_mct (in atm_comp_mct.F90) is added, recording information for all MPI tasks and in a more compact representation for each CAM instance. b) Output CLM MPI task to compute node mapping using share routine MPI task to compute node mapping information in CLM is output one line per task, and only for the first 100 MPI tasks. A new share routine in CIME (shr_taskmap_write) is available that writes out one line per compute node. Each line contains the compute node name and the list of MPI tasks assigned to that node for a given communicator. Here the CLM logic in spmd_init (in spmdMod.F90) is removed and a call to shr_taskmap_write in lnd_init_mct (in lnd_comp_mct.F90) is added, recording information for all MPI tasks and in a more compact representation for each CLM instance. c) Output Mosart MPI task to compute node mapping using share routine MPI task to node mapping information is output for ROF instances using the new share routine shr_taskmap_write. The call is added to the routine rof_init_mct in rof_comp_mct.F90. d) Output mpas-albany-landice MPI task to compute node mapping MPI task to node mapping information is output for mpas-albany-landglc instances using the new share routine shr_taskmap_write. The call is added to the routine glc_init_mct in glc_comp_mct.F. e) Output mpas-ocean MPI task to compute node mapping using share routine MPI task to node mapping information is output for mpas-ocean instances using the new share routine shr_taskmap_write. The call is added to the routine ocn_init_mct in ocn_comp_mct.F. f) Output mpas-seaice MPI task to compute node mapping using share routine MPI task to node mapping information is output for mpas-seaice instances using the new share routine shr_taskmap_write. The call is added to the routine ice_init_mct in ice_comp_mct.F. Fixes #2485 [BFB] 26 November 2018, 22:52:07 UTC
4153946 mark-petersen 26 November 2018, 22:23:34 UTC Update mpas-source submodule 26 November 2018, 22:23:34 UTC
6bd3ec1 Patrick Worley 26 November 2018, 17:15:42 UTC Update system-specific job progress monitoring scripts When developing the job progress monitoring script for Summit, some improvements were identified that apply equally to the scripts for the other systems (syslog.<system-name>). On Summit, background processes ignore kill signals while they are asleep, and job scripts will not complete until spawned background processes are killed or complete normally. Thus, if the job progress monitoring script sleeps for a long time (a user-defined interval), allocation time will be wasted. The workaround was to divide the sleep interval into a loop over 2 minute sleep intervals. This same change of implementation is implemented here for all systems, to make sure that background jobs do not outlive the spawning job scripts. Currently, some system status information is collected twice (with output of the second call overwriting that of the first). A second clean-up is to remove this redundancy. The e3sm log file is copied to the checkpoint directory at the beginning of the run, in an effort to capture the task-to-node mapping. However, this copy needs to be delayed long enough for the mapping information to show up in the log file first. Various methods have been tried to calculate the number of lines in e3sm.log that would indicate that this information has been captured. Yet another approach is implemented here, using empirically determined system-specific and job-specific line counts. (Note that we cannot finish the update for Summit until Summit is available for testing again.) Finally, since Mira and Cetus are no longer target systems for the E3SM project, the corresponding syslog scripts are deleted. [BFB] 26 November 2018, 17:15:42 UTC
86bed07 Jon Wolfe 21 November 2018, 18:44:02 UTC Update the MOSART 1/8th degree global parameter file 21 November 2018, 18:44:02 UTC
fc73818 James Foucar 15 November 2018, 17:27:45 UTC Merge branch 'jgfouca/cime/add_latband' into master (PR #2619) Add optional stream for co2 timeseries with latitude bands to datm. [BFB] * origin/jgfouca/cime/add_latband: Add E3SM support for latband datm Update CIME to ESMCI cime5.7.3-2 (PR #2599) 19 November 2018, 21:34:08 UTC
b0927e6 Patrick Worley 19 November 2018, 16:27:15 UTC Change write format specification to match existing style Modified write format specification to match the style used elsewhere in the same subroutines. [BFB] 19 November 2018, 16:27:15 UTC
ac53159 Junqi Yin 19 November 2018, 03:30:07 UTC Merge branch 'bishtgautam/lnd/rename-bgc-files' (PR #2613) Renames land BGC files Renames land biogeochemistry files to follow a consistent naming convention. [BFB] 19 November 2018, 03:30:07 UTC
2ff7e0c Balwinder Singh 16 November 2018, 15:50:31 UTC Modifies lmod path [BFB] - Bit-For-Bit 16 November 2018, 15:50:31 UTC
067f39e James Foucar 15 November 2018, 17:48:29 UTC Merge branch 'jgfouca/cime/fix_configure' into master (PR #2626) Fix for cime's configure tool Needed comp_interface. [BFB] * jgfouca/cime/fix_configure: Fix for cime's configure tool 15 November 2018, 17:49:11 UTC
bcf6f74 James Foucar 15 November 2018, 17:37:35 UTC Fix for cime's configure tool Needed comp_interface. [BFB] 15 November 2018, 17:37:35 UTC
e95e91c Patrick Worley 15 November 2018, 15:06:21 UTC Add control of taskmap output to env_run.xml Setting the new namelist parameters is moved to env_run.xml, with the same defaults for E3SM but both disabled for CESM. [BFB] 15 November 2018, 15:06:21 UTC
0e45723 Junqi Yin 15 November 2018, 14:45:18 UTC Merge branch 'bishtgautam/lnd/water-balance-master' (PR #2603) Adds water budget diagnostics in the land model Following improvements to the water budget in the land model are added: 1. Fixes the incorrect update of water state of unconfined aquifer at the beginning of the land driver. This fix improves the net water budget in the fully coupled simulation. 2. Calculation of a grid-level water budget is now included. 3. Diagnostics about water budget can be now outputted in the land logfiles by setting do_budgets=.true. in usr_nl_clm. The diagnostics includes: a. Temporally integrated water fluxes over a given period (e.g. instantaneus, daily, monthly, annual, all-time). This diagnostic information is similar to the information outputted in the coupler logfile if BUDGETS=TRUE in env_run.xml. b. Water states at the beginning and end of the integration period. This diagnostic information includes water stored in various land components (i.e. soil moisture, canopy storage, etc). 4. Additional data is saved in land monthly history tapes to recreate the land water diagnostics from model output. 5. Additional data is saved in land restart file to support output of land water diagnostics when a simulation restarts in the middle of a month. Fixes #2605 [non-BFB] 15 November 2018, 14:45:18 UTC
b5e4789 Oksana Guba 15 November 2018, 03:43:56 UTC minor fix 15 November 2018, 03:43:56 UTC
e204866 Patrick Worley 15 November 2018, 01:29:17 UTC Control ouput of MPI task to compute node mapping data A recent PR fixed the output of component-specific MPI task to compute node mapping for ATM and LND, and added it for ROF, OCN, ICE, and GLC. The utility of this output for each component is unclear, and a namelist variable to enable or disable it is added here. This variable also allows setting the format of the output. All component model task-to-node mapping output is controlled by this single cime_pes namelist variable info_taskmap_comp, where setting this to 0 disables output, setting it to 1 uses the more compact format, and setting it to 2 uses the more verbose format. Note that a separate component-only PR will be needed for the components to be aware of info_taskmap_comp and use this as described above, but this additional code has been used to test this PR. For symmetry, a cime_pes namelist variable info_taskmap_model was also added to control the task-to-node mapping output for the full coupled model, with the same options (0, 1, and 2). The default here is 2 (preserving the current behavior). Also for symmetry, output of task-to-node mapping data was added for the CPL component (and is controlled by info_taskmap_comp the same as for the other component models). Finally, the implementation for the whole-model task-to-node mapping data output was modified to better reflect the (currently incomplete?) support for running the model with multiple drivers. [BFB] [NML] 15 November 2018, 01:29:17 UTC
25ed9e2 Jon Wolfe 14 November 2018, 15:48:27 UTC Merge branch 'jonbob/scripts/add_E3SM-ARRM' (PR #2616) Add E3SM-ARRM (Arctic refined) grid configurations This PR adds the MPAS-O and MPAS-SI meshes and other configuration files necessary to run a G-case of two E3SM-ARRM (Arctic refined) configurations: * the first, lower resolution grid, is EC60to30 in the Southern * hemisphere and then transitions from 30 to 10 km in the Northern hemisphere (10km resolution over the whole Arctic); * the second, higher resolution grid, is similar to the one in 1), but transitions to 6 km in the Arctic. Testing has been done on the LANL machine grizzly and on edison. One test ran successfully for 10 years. [BFB] 14 November 2018, 15:48:27 UTC
2b52b03 Gautam Bisht 13 November 2018, 17:51:08 UTC Merge branch 'yangx/lnd/bugfix_C13_C14' (PR #2610) Fixed a few bugs in the history file module - several C14 and C13 variables were added twice there, which caused the model to crash when C13 and C14 are turned on. Fixes #2611 [BFB] 13 November 2018, 17:51:08 UTC
e9f2f04 Andrew M. Bradley 12 November 2018, 22:32:36 UTC Address the SFC part of scalably initializing mesh-file-based grid for RRM. Previously, mesh_mod::initialize_space_filling_curve used an unscalable method to find space-filling curve indices for the GridVertex array. This commit implements a new method that uses the new spacecurve::sfcmap_pos2i function to obtain the indices using O(log ne2) space instead of O(ne2^2) space and O(nelem log(ne2)) time instead of O(ne2^2) time. [BFB] 13 November 2018, 00:48:57 UTC
f93efd2 Oksana Guba 13 November 2018, 00:42:07 UTC cleaning up timelevels, fixing the same ps bug in theta model (though not really tested) 13 November 2018, 00:42:07 UTC
6a74539 Oksana Guba 12 November 2018, 23:20:02 UTC fixing a bug in conversion routine for dry cases, ps at the next time level wasnt init-ed properly 12 November 2018, 23:20:02 UTC
60e10ba James Foucar 12 November 2018, 23:06:30 UTC Add E3SM support for latband datm [BFB] 12 November 2018, 23:06:30 UTC
b50d979 Andrew M. Bradley 12 November 2018, 20:40:47 UTC Merge remote-tracking branch 'upstream/master' into ambrad/homme/scalable-init-phase2a 12 November 2018, 20:40:47 UTC
343121e Gautam Bisht 12 November 2018, 17:47:39 UTC Renames CNSharedParamsMod.F90 12 November 2018, 17:47:39 UTC
72c2867 Ryan Knox 12 November 2018, 01:37:30 UTC Updated the FATES repository to use NGEET/ 12 November 2018, 01:37:30 UTC
ca5c279 Gautam Bisht 10 November 2018, 18:16:55 UTC Renames CNVerticalProfileMod.F90 10 November 2018, 18:16:55 UTC
6d58c86 Gautam Bisht 10 November 2018, 18:11:55 UTC Renames CNVegStructUpdateMod.F90 10 November 2018, 18:11:55 UTC
74f1fab Gautam Bisht 10 November 2018, 18:09:03 UTC Renames CNRootDynMod.F90 10 November 2018, 18:09:03 UTC
987f84f Gautam Bisht 10 November 2018, 17:45:16 UTC Renames CNMRespMod.F90 10 November 2018, 17:45:16 UTC
back to top